Mova0.5 pre-release⌕
☰
home / corpora / questions / AI / What is AI? Pattern-guessers, not people / e4

Exam question 4 · What is AI? Pattern-guessers, not people

An image classifier is observed to designate a fluffy slipper as a cat. Ascertain the inference which may properly be drawn.

Mova's choice: option 1 · The right answer: option 1 · right ✓ · Who acts: you · What happens: observe, designate, ascertain, draw · To what, where: image classifier, fluffy slipper, inference

This is exam question 4 of lesson 1 «What is AI? Pattern-guessers, not people» of the AI course of Mova's primary school, asked in stiff, formal language. The page gives the question's syntax tree — a dependency tree in the style of Universal Dependencies — and its pragmatics clause by clause, then the 4 options as the student saw them, each with its tree and pragmatics. Every text is parsed by Mova itself. How the student answered it, with the score of every option and why: the test at school.

The question

Sentence 1: An image classifier is observed to designate a fluffy slipper as a cat.

detcompoundnsubj:passaux:passmarkxcompdetamodobjcasedetoblpunctrootAnDETimageNOUNclassifierNOUNisAUXobservedVERBtoPARTdesignateVERBaDETfluffyADJslipperNOUNasADPaDETcatNOUN.PUNCT
The words, one row each (CoNLL-U)
#wordlemmaUPOSfeaturesheadrelation
1AnaDETDefinite=Ind PronType=Art3 classifierdet
2imageimageNOUNNumber=Sing3 classifiercompound
3classifierclassifierNOUNNumber=Sing5 observednsubj:pass
4isbeAUXMood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Fin5 observedaux:pass
5observedobserveVERBTense=Past VerbForm=Part Voice=Passrootroot
6totoPART_7 designatemark
7designatedesignateVERBVerbForm=Inf5 observedxcomp
8aaDETDefinite=Ind PronType=Art10 slipperdet
9fluffyfluffyADJDegree=Pos10 slipperamod
10slipperslipperNOUNNumber=Sing7 designateobj
11asasADP_13 catcase
12aaDETDefinite=Ind PronType=Art13 catdet
13catcatNOUNNumber=Sing7 designateobl
14..PUNCT_5 observedpunct
#verbtense, moodwhowhat, whomto whomwherewhenotherroles by
1observepresent · passivea image classifiertree rules
2designateto-infinitivea fluffy slipperas a catvn appoint-29.1 «NP V NP PP.attribute»: agent=Agent slipper=Theme as cat=Attribute

One row per clause the reader finds. «not said here»: a role filled from the clause above. «roles by»: a VerbNet class and frame when one fits the verb, else the reader's rules on the tree.

Sentence 2: Ascertain the inference which may properly be drawn.

detobjnsubj:passauxadvmodaux:passacl:relclpunctrootAscertainVERBtheDETinferenceNOUNwhichPRONmayAUXproperlyADVbeAUXdrawnVERB.PUNCT
The words, one row each (CoNLL-U)
#wordlemmaUPOSfeaturesheadrelation
1AscertainascertainVERBMood=Imp VerbForm=Finrootroot
2thetheDETDefinite=Def PronType=Art3 inferencedet
3inferenceinferenceNOUNNumber=Sing1 Ascertainobj
4whichwhichPRONPronType=Rel8 drawnnsubj:pass
5maymayAUXVerbForm=Fin8 drawnaux
6properlyproperlyADV_8 drawnadvmod
7bebeAUXVerbForm=Inf8 drawnaux:pass
8drawndrawVERBTense=Past VerbForm=Part Voice=Pass3 inferenceacl:relcl
9..PUNCT_1 Ascertainpunct
#verbtense, moodwhowhat, whomto whomwherewhenotherroles by
1ascertainimperativeyou not said herethe inferencevn discover-84 «NP V NP»: agent=Agent inference=Topic
2drawmay · passivethe inferenceproperly propertyvn scribble-25.2-1 «NP V NP»: subject=Theme

The options

Option 1 · ← Mova's choice · ✓ the right answer

The classifier has learned visual resemblance from examples and may be misled by similar objects

The tree, 15 words
detnsubjauxamodobjcaseoblccauxaux:passconjcaseamodoblrootTheDETclassifierNOUNhasAUXlearnedVERBvisualADJresemblanceNOUNfromADPexamplesNOUNandCCONJmayAUXbeAUXmisledVERBbyADPsimilarADJobjectsNOUN
The words, one row each (CoNLL-U)
#wordlemmaUPOSfeaturesheadrelation
1ThetheDETDefinite=Def PronType=Art2 classifierdet
2classifierclassifierNOUNNumber=Sing4 learnednsubj
3hashaveAUXMood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Fin4 learnedaux
4learnedlearnVERBTense=Past VerbForm=Partrootroot
5visualvisualADJDegree=Pos6 resemblanceamod
6resemblanceresemblanceNOUNNumber=Sing4 learnedobj
7fromfromADP_8 examplescase
8examplesexampleNOUNNumber=Plur4 learnedobl
9andandCCONJ_12 misledcc
10maymayAUXVerbForm=Fin12 misledaux
11bebeAUXVerbForm=Inf12 misledaux:pass
12misledmisleadVERBTense=Past VerbForm=Part Voice=Pass4 learnedconj
13bybyADP_15 objectscase
14similarsimilarADJDegree=Pos15 objectsamod
15objectsobjectNOUNNumber=Plur12 misledobl
#verbtense, moodwhowhat, whomto whomwherewhenotherroles by
1learnpresent · perfectthe classifiervisual resemblanceexamplesvn learn-14-2-1 «NP V NP PP.source»: agent=Agent resemblance=Topic from example=Source
2misleadmay · passivesimilar objectsvn trick-59.2 «NP V NP»: agent=Agent

Option 2

The classifier possesses knowledge that the slipper belongs to a cat

detnsubjobjmarkdetnsubjadvclcasedetoblrootTheDETclassifierNOUNpossessesVERBknowledgeNOUNthatSCONJtheDETslipperNOUNbelongsVERBtoADPaDETcatNOUN
The words, one row each (CoNLL-U)
#wordlemmaUPOSfeaturesheadrelation
1ThetheDETDefinite=Def PronType=Art2 classifierdet
2classifierclassifierNOUNNumber=Sing3 possessesnsubj
3possessespossessVERBMood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Finrootroot
4knowledgeknowledgeNOUNNumber=Sing3 possessesobj
5thatthatSCONJ_8 belongsmark
6thetheDETDefinite=Def PronType=Art7 slipperdet
7slipperslipperNOUNNumber=Sing8 belongsnsubj
8belongsbelongVERBMood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Fin3 possessesadvcl
9totoADP_11 catcase
10aaDETDefinite=Ind PronType=Art11 catdet
11catcatNOUNNumber=Sing8 belongsobl
#verbtense, moodwhowhat, whomto whomwherewhenotherroles by
1possesspresentthe classifierknowledgevn own-100.1 «NP V NP»: agent=Pivot knowledge=Theme
2belongpresentthe slipperto a catvn attend-107.4-2 «NP V PP»: agent=Theme

Option 3

The classifier has been deliberately instructed to deceive its users

detnsubj:passauxaux:passadvmodmarkxcompnmod:possobjrootTheDETclassifierNOUNhasAUXbeenAUXdeliberatelyADVinstructedVERBtoPARTdeceiveVERBitsPRONusersNOUN
The words, one row each (CoNLL-U)
#wordlemmaUPOSfeaturesheadrelation
1ThetheDETDefinite=Def PronType=Art2 classifierdet
2classifierclassifierNOUNNumber=Sing6 instructednsubj:pass
3hashaveAUXMood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Fin6 instructedaux
4beenbeAUXTense=Past VerbForm=Part6 instructedaux:pass
5deliberatelydeliberatelyADV_6 instructedadvmod
6instructedinstructVERBTense=Past VerbForm=Part Voice=Passrootroot
7totoPART_8 deceivemark
8deceivedeceiveVERBVerbForm=Inf6 instructedxcomp
9itsitsPRONCase=Gen Gender=Neut Number=Sing Person=3 Poss=Yes PronType=Prs10 usersnmod:poss
10usersuserNOUNNumber=Plur8 deceiveobj
#verbtense, moodwhowhat, whomto whomwherewhenotherroles by
1instruct speech: otherpresent · perfect · passivethe classifierdeliberately propertyvn advise-37.9-1 «NP V NP how S_INF»: subject=Recipient clause=Topic
2deceiveto-infinitiveclassifier's usersvn trick-59.2 «NP V NP»: agent=Agent user=Patient

Option 4

The classifier is functioning without error, slippers being a variety of cat

detnsubjcopamodcaseoblpunctconjcopdetcasenmodrootTheDETclassifierNOUNisAUXfunctioningADJwithoutADPerrorNOUN,PUNCTslippersVERBbeingAUXaDETvarietyNOUNofADPcatNOUN
The words, one row each (CoNLL-U)
#wordlemmaUPOSfeaturesheadrelation
1ThetheDETDefinite=Def PronType=Art2 classifierdet
2classifierclassifierNOUNNumber=Sing4 functioningnsubj
3isbeAUXMood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Fin4 functioningcop
4functioningfunctioningADJDegree=Pos11 varietyamod
5withoutwithoutADP_6 errorcase
6errorerrorNOUNNumber=Sing4 functioningobl
7,,PUNCT_8 slipperspunct
8slippersslipperVERBMood=Ind Number=Sing Person=3 Tense=Pres VerbForm=Fin4 functioningconj
9beingbeAUXTense=Pres VerbForm=Part11 varietycop
10aaDETDefinite=Ind PronType=Art11 varietydet
11varietyvarietyNOUNNumber=Singrootroot
12ofofADP_13 catcase
13catcatNOUNNumber=Sing11 varietynmod
#verbtense, moodwhowhat, whomto whomwherewhenotherroles by
1ispresentthe classifierwithout error; functioning propertytree rules
2slipperpresentthe classifier not said heretree rules
3is agerunda functioning varietyof cattree rules

Text: the tests of Mova's primary school, written for Mova (world/data/school/), as the course runner gave them to the student. Trees: Mova's own parse — the en model on the CPU, when the site is built — not checked by a teacher; they may be wrong. Frames: Mova's frames reader (prag), each text read on its own. Elamur Cosmic License.

This page in Markdown → https://md.mova.elamur.ai/free/en/corpora/questions/ai/1-what-is-ai/e4/